The generation of protein-protein interaction network (interactome) models with increasing quality and sensitivity is a necessary, although not sufficient, aspect in the quest of generating predictive systems models. The yeast S. cerevisiae is an excellent model to validate this concept. Two major approaches, yielding complementary high-quality (HQ) data are used to experimentally map interactomes: i) binary physical interaction mapping; and ii) and protein complex analysis. Both approaches are required to obtain a complete view of the interactome of an organism. We recently developed a conceptual framework to assess interactome models based on quantitative benchmarking of screening and validation assays against reference sets. The framework then combines specific measurements of completeness, assay sensitivity, sampling sensitivity and precision to estimate the overall sensitivity and quality of a network model. With this we have shown that early yeast interactome data cover ~10%, our second generation map ~20%, of the yeast binary interactome with HQ interactions. For increasingly standardized and improved quality control, we developed an experimental method to assign confidence scores to individual interactions. This method uses a tool-kit of interaction assays each benchmarked against common reference sets. Experimental validation of every interaction in the tool-kit assays enables integration of benchmark and validation data to calculate individual confidence scores. Here we propose to continue the mapping efforts for the binary interactome of S. cerevisiae. Utilizing novel technologies we aim to extend the overall sensitivity to 50%, which is both technically feasible and can be expected to enable a profoundly improved understanding of the interactome network and how it mediates genotype-to-phenotype relationships. For quality control all interactions will be validated in multiple standardized binary interaction assays and a confidence score for each individual interaction will be calculated. We will thus generate and analyze a third generation binary interactome map of S. cerevisiae. Our specific aims are: i) to expand the binary interactome map of S. cerevisiae from ~20 to ~50% sensitivity, ii) to validate experimentally all binary interactions found in Specific Aim 1 with a novel confidence scoring strategy based on a highly-controlled and benchmarked assay tool-kit. iii) To expand the global analysis of the S. cerevisiae binary interactome network using the new high- quality map obtained in Specific Aim 2.